ARTICLE IN PRESS Quaternary Science Reviews 27 (2008) 882–887 Why there was not a Younger Dryas-like event during the Penultimate Deglaciation Anders E. Carlsonà Department of Geology and Geophysics, University of Wisconsin, 1215 West Dayton Street, Madison, WI 53706, USA Received 16 August 2007; received in revised form 14 February 2008; accepted 15 February 2008 Abstract The Younger Dryas cold event is a relatively unique feature of the last deglaciation when compared to previous deglaciations, suggesting a unique trigger rather than the commonly held forcing mechanism of North American freshwater routing to the North Atlantic. Here, I compare the last (T-I) and penultimate (T-II) deglaciations and provide new support for the argument that the lack of a Younger Dryas-like event during T-II is due to the rapidity of Northern Hemisphere ice sheet retreat under greater boreal summer insolation forcing. Faster ice retreat suppressed Atlantic meridional overturning circulation (AMOC) until near the end of T-II, while during T-I AMOC increased relatively early. During T-I, the eastward routing of freshwater that caused the Younger Dryas happened after AMOC resumption, whereas during T-II this routing occurred prior to the resumption of AMOC. Thus the increased flux of freshwater to the North Atlantic during T-II had little effect on AMOC, explaining the lack of a Younger Dryas-like climate oscillation during this deglaciation. r 2008 Elsevier Ltd. All rights reserved. 1. Introduction tic, reducing Atlantic meridional overturning circulation (AMOC) and northward heat transport (McManus et al., The Younger Dryas cold event (12.9–11.5 kyr BP) 2004)(Fig. 1a). Subsequent paleoceanographic studies interrupted the last deglaciation (19–6 kyr BP) with an questioned this mechanism because of the lack of evidence abrupt return to cold, glacial conditions (Fig. 1a). This for St. Lawrence Estuary freshening during the Younger event is recorded in numerous records from around the Dryas (Keigwin and Jones, 1995; deVernal et al., 1996), North Atlantic region (Clark et al., 2002), in subtropical which in part led modeling studies to suggest alternative records influenced by monsoon intensity (Yuan et al., 2004; freshwater sources as the cause of this event (Tarasov and Kelly et al., 2006; Weldeab et al., 2007), and in records of Peltier, 2005). Furthermore, minimum limiting radiocar- atmospheric methane concentration mainly reflecting pre- bon dates from the western Lake Superior region have been cipitation and temperature impacts on boreal and tropical interpreted as indicating that ice retreated out of Lake wetlands (Brook et al., 2000; Monnin et al., 2001; Schaefer Superior after the onset of the Younger Dryas (Lowell et al., 2006; Sowers, 2006)(Fig. 1b and c). The Younger et al., 2005; Teller et al., 2005), but these dates are only Dryas was originally hypothesized to be caused by the minimum limiting and thus do not refute the original retreat of the southern Laurentide Ice Sheet (LIS) out of routing hypothesis (Teller and Boyd, 2006). Recently, Lake Superior with attendant routing of western Canadian geochemical records from the St. Lawrence Estuary Plains freshwater from the Mississippi River to the St. contested these studies and likely confirmed the original Lawrence River 12.9 kyr BP (Johnson and McClure, routing hypothesis as the cause of the Younger Dryas 1976; Rooth, 1982; Broecker et al., 1989). This increased (Carlson et al., 2007). However, evidence for a Younger freshwater discharge freshened the northern North Atlan- Dryas-like event is equivocal during earlier deglaciations (Sarnthein and Tiedmann, 1990; Adkins et al., 1997; ÃCorresponding author. Tel.: +1 608 262 1921; fax: +1 608 262 0693. Oppo et al., 1997, 2001; Chapman and Shackleton, 1998; E-mail address: [email protected] Petit et al., 1999; Spahni et al., 2005; Kelly et al., 2006; 0277-3791/$ - see front matter r 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.quascirev.2008.02.004 ARTICLE IN PRESS A.E. Carlson / Quaternary Science Reviews 27 (2008) 882–887 883 Fig. 1. Comparing North Atlantic influenced proxies during T-I (left) and T-II (right). (a) GISP2 d18O, a proxy of Summit Greenland temperature (Grootes et al., 1993). (b) T-I Dome C (Monnin et al., 2001) and T-II Vostok (Petit et al., 1999) atmospheric methane concentration (black line with round symbols). Also shown during T-II is the alternate age model for methane of Ruddiman and Raymo (2003) (thin black line). The Shackleton (2000) age model falls between these two. (c) Dongge Cave speleothem d18O(Yuan et al., 2004; Kelly et al., 2006), a proxy for SEAM intensity. Note that these records are similar to proxy records of West African monsoon intensity (Weldeab et al., 2007). (d) Relative sea level; T-I is from the Clark and Mix (2002) compilation, and T-II is the revised coral dating of Thompson and Goldstein (2005) with their interpreted sea level curve (dashed black lines). Note that the magnitude of the T-II record is in agreement with Red Sea records (Siddall et al., 2006; Rohling et al., 2008). (e) June insolation at 601 N(Berger and Loutre, 1991). Dark gray bars denote the timing of red bed deposition in the Labrador Sea. Medium gray bar during T-I is the timing of the Younger Dryas (YD) cold period. Light gray bar is the timing of the Bølling/Allerød (B/A) warm period. Dashed vertical line during T-II denotes the timing of increased SEAM intensity and the inferred turn-on of AMOC, similar to the onset of the Bølling during T-I. Desprat et al., 2007; Weldeab et al., 2007), implying warm period (B/A). This abrupt warming is also expressed as a unique trigger for the event rather than the routing an intensification of the Southeast Asian monsoon (SEAM) mechanism (Firestone et al., 2007), which would presum- (Yuan et al., 2004)(Fig. 1c) and increased atmospheric ably operate during every deglaciation (Broecker, 2006). methane concentration (Fig. 1b) reflecting greater production Here, I synthesize key deglacial records along with new in tropical and boreal wetlands (Brook et al., 2000). The evidence that support the hypothesis that the lack of a subsequent routing of freshwater to the St. Lawrence River Younger Dryas-like event during the penultimate deglacia- reduced AMOC 12.9 kyr BP with attendant cooling of the tion (Termination II; T-II) is due to the magnitude of North Atlantic region, diminished monsoon activity and deglacial forcing and the resulting climate setting relative decreased wetland productivity (Fig. 1)(Brook et al., 2000; to the last deglaciation (Termination I; T-I), not a unique Clark et al., 2002; Yuan et al., 2004; Carlson et al., 2007). By trigger. the beginning of the Younger Dryas, the LIS had lost 1/3 of its overall LGM extent (Dyke, 2004), contributing 20 m 2. Deglacial events during T-I and T-II to the 60 m of deglacial sea level rise (Fig. 1d). Note that the Younger Dryas and 1/2 of the overall deglacial sea The major warming of the North Atlantic region following level rise occurred near the peak in boreal summer insolation the last glacial maximum (LGM; 21 kyr BP) occurred (Fig. 1d and e). 14.7 kyr BP due to an increase in AMOC (Fig. 1a) During T-II, paleoceanographic data suggest a short- (McManus et al., 2004) and is called the Bølling/Allerød lived deglacial cooling event of a smaller magnitude than ARTICLE IN PRESS 884 A.E. Carlson / Quaternary Science Reviews 27 (2008) 882–887 the Younger Dryas that is restricted to the eastern North The contrasting relative timings of deglacial AMOC Atlantic, possibly in response to variations in shallow resumption during T-I and T-II may be in response to the North Atlantic overturning (Oppo et al., 2001; Desprat greater boreal summer insolation increase that forced T-II et al., 2007). This local climate oscillation is not reflected in relative to T-I (Fig. 1 and Table 1)(Ruddiman et al., 1980; deeper AMOC records (Sarnthein and Tiedmann, 1990; Oppo et al., 1997). During T-II, global ice sheet retreat Adkins et al., 1997; Oppo et al., 1997, 2001; Chapman and lasted 14 kyr from the initial increase in boreal summer Shackleton, 1998), or records of atmospheric methane insolation to the achievement of full deglaciation2 (0 m of concentration and SEAM intensity (Petit et al., 1999; Kelly relative sea level), relative to 19 kyr during T-I (Fig. 1). et al., 2006), suggesting that a large increase and The effects of greater insolation forcing are also evident in subsequent decrease in AMOC did not occur during T-II the time the LIS took to retreat to a greatly diminished size. (Fig. 1). Rather, atmospheric methane and SEAM abruptly One indicator of a greatly diminished LIS is the deposition increased 129 kyr BP1 similar to the B/A onset during T-I of red clay sediment from northern Hudson Bay in the (Fig. 1). Given the correlation between AMOC, SEAM, Labrador Sea, which indicates the collapse of ice over and atmospheric methane (Brook et al., 2000; McManus Hudson Bay and attendant eastward drainage of north- et al., 2004; Yuan et al., 2004; Broecker, 2006), the SEAM- western Hudson Bay water (Barber et al., 1999). During methane increases at 129 kyr PB imply a coeval abrupt T-I, the deposition of this red bed in the Labrador Sea increase in AMOC. Similarly, Oppo et al. (1997) demon- occurred 8.4 kyr BP (Barber et al., 1999) when sea level strated that North Atlantic benthic d13C records suggest a was À15 m (Fig. 1). During T-II, a similar carbonate red resumption of AMOC at this abrupt transition.